Analysis of the algorithm: From embeddings to prioritized genes.

The algorithm transformed the similarity matrix to make it compatible with the embedding process. Once this was done for each network and embedding type, it was integrated by embedding type. Below there is a general analysis of the properties of each matrix in the different phases of the process, including the graph building process for each layer.

Annotations Properties

Table 1. Annotation descriptors.

Net Min Max Average Standard_Deviation
DepMap_effect 972 17354 2043.6562906724512 3829.1218209282715
biological_process 1 1082 8.241812188793043 20.853311380298706
cellular_component 1 5172 7.219656420451649 82.56741928666138
disease 1 81 1.7829472267615112 2.3512948371338793
gene_PS 1 108 2.1519058295964126 5.045578888786698
gene_TF 1 5778 3.0320060963993143 74.46796191948712
gene_hgncGroup 1 2294 2.261042369703787 23.27548667222241
hippie_ppi 1 2576 45.37051659901017 93.00135289510746
molecular_function 1 6791 4.87167748867193 54.08093332133869
pathway 1 479 6.75436035343834 16.179001796162975
phenotype 1 1307 23.935561037601854 48.703049458006625
string_ppi_coexpression 1 4044 339.28043549983505 359.91355895653743
string_ppi_combined_score 1 7338 610.0699285559645 500.6662722348227
string_ppi_cooccurence 1 133 20.415675297410775 27.990489848424655
string_ppi_database 1 1287 64.01997573042098 86.01865153961889
string_ppi_exp 1 4192 249.92115027829314 354.50679835712316
string_ppi_experimental 1 4192 249.92115027829314 354.50679835712316
string_ppi_fusion 1 101 3.721608040201005 7.370788836174388
string_ppi_neighborhood 1 1391 92.45592392701106 127.04630294501078
string_ppi_textmining 1 7313 521.6452470039586 444.68260699568225

Individual Processing Graph steps

<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<

Embedding Process

Table 2. Uncombined Embedding Matrixes

Net Kernel Matrix_Dimensions Matrix_Elements Matrix_Elements_Non_Zero Matrix_Non_Zero_Density
DepMap_effect_pearson ct 17354x17354 301161316 301161316 1.0
DepMap_effect_pearson el 17354x17354 301161316 301161316 1.0
DepMap_effect_pearson ka 17354x17354 301161316 106778670 0.35455639329189276
DepMap_effect_pearson node2vec 17354x17354 301161316 301161314 0.9999999933590409
DepMap_effect_pearson rf 17354x17354 301161316 301161316 1.0
DepMap_effect_spearman ct 17354x17354 301161316 301161316 1.0
DepMap_effect_spearman el 17354x17354 301161316 301161316 1.0
DepMap_effect_spearman ka 17354x17354 301161316 194719108 0.6465608219084817
DepMap_effect_spearman node2vec 17354x17354 301161316 301161316 1.0
DepMap_effect_spearman rf 17354x17354 301161316 301161316 1.0
biological_process ct 16978x16978 288252484 288252484 1.0
biological_process el 16978x16978 288252484 288252484 1.0
biological_process ka 16978x16978 288252484 249741002 0.8663967038008248
biological_process node2vec 16978x16978 288252484 288252484 1.0
biological_process raw_sim 16978x16978 288252484 249741002 0.8663967038008248
biological_process rf 16978x16978 288252484 288252484 1.0
cellular_component ct 17963x17963 322669369 322669369 1.0
cellular_component el 17963x17963 322669369 322669369 1.0
cellular_component ka 17963x17963 322669369 322669369 1.0
cellular_component node2vec 17963x17963 322669369 322669369 1.0
cellular_component rf 17963x17963 322669369 322669369 1.0
disease ct 3929x3929 15437041 15437041 1.0
disease el 3929x3929 15437041 15437041 1.0
disease ka 3929x3929 15437041 15325785 0.992792919316597
disease node2vec 3929x3929 15437041 15437041 1.0
disease raw_sim 3929x3929 15437041 15325785 0.992792919316597
disease rf 3929x3929 15437041 15437041 1.0
gene_PS ct 3020x3020 9120400 8927292 0.978826805841849
gene_PS el 3020x3020 9120400 5842664 0.6406148853120477
gene_PS ka 3020x3020 9120400 98718 0.010823867374237971
gene_PS node2vec 3020x3020 9120400 9120400 1.0
gene_PS rf 3020x3020 9120400 5842664 0.6406148853120477
gene_TF ct 3871x3871 14984641 14984637 0.9999997330600046
gene_TF el 3871x3871 14984641 14969165 0.9989672091576969
gene_TF ka 3871x3871 14984641 5275373 0.35205201112258877
gene_TF node2vec 3871x3871 14984641 14984641 1.0
gene_TF rf 3871x3871 14984641 14969165 0.9989672091576969
gene_hgncGroup ct 25136x25136 631818496 631304330 0.9991862124909999
gene_hgncGroup el 25136x25136 631818496 326932198 0.5174463870079549
gene_hgncGroup ka 25136x25136 631818496 14482790 0.022922390040319426
gene_hgncGroup node2vec 25136x25136 631818496 631818496 1.0
gene_hgncGroup raw_sim 25136x25136 631818496 14457654 0.022882606462980154
gene_hgncGroup rf 25136x25136 631818496 326932198 0.5174463870079549
hippie_ppi ct 13444x13444 180741136 180714249 0.9998512402843368
hippie_ppi el 13444x13444 180741136 175403916 0.9704703637582537
hippie_ppi ka 13444x13444 180741136 213164 0.0011793884044194567
hippie_ppi node2vec 15763x15763 248472169 248472167 0.9999999919508088
hippie_ppi rf 13444x13444 180741136 175403916 0.9704703637582537
molecular_function ct 17333x17333 300432889 300432889 1.0
molecular_function el 17333x17333 300432889 300432889 1.0
molecular_function ka 17333x17333 300432889 300432889 1.0
molecular_function node2vec 17333x17333 300432889 300432889 1.0
molecular_function rf 17333x17333 300432889 300432889 1.0
pathway ct 5611x5611 31483321 31482925 0.9999874219114305
pathway el 5611x5611 31483321 25851467 0.8211162666098663
pathway ka 5611x5611 31483321 362997 0.011529819233491919
pathway node2vec 5611x5611 31483321 31483321 1.0
pathway rf 5611x5611 31483321 25851467 0.8211162666098663
phenotype ct 5074x5074 25745476 25745476 1.0
phenotype el 5074x5074 25745476 25745476 1.0
phenotype ka 5074x5074 25745476 25745476 1.0
phenotype node2vec 5074x5074 25745476 25745476 1.0
phenotype raw_sim 5074x5074 25745476 25745476 1.0
phenotype rf 5074x5074 25745476 25745476 1.0
string_ppi_coexpression ct 18186x18186 330730596 330730596 1.0
string_ppi_coexpression el 18186x18186 330730596 330148940 0.9982412996951755
string_ppi_coexpression ka 18186x18186 330730596 6188334 0.018711102253146244
string_ppi_coexpression node2vec 18186x18186 330730596 330730596 1.0
string_ppi_coexpression raw_sim 18186x18186 330730596 6170151 0.01865612397106435
string_ppi_coexpression rf 18186x18186 330730596 330148940 0.9982412996951755
string_ppi_cooccurence ct 2858x2858 8168164 8167760 0.9999505396806431
string_ppi_cooccurence el 2858x2858 8168164 1409490 0.17255897408524118
string_ppi_cooccurence ka 2858x2858 8168164 61198 0.007492258970314504
string_ppi_cooccurence node2vec 2858x2858 8168164 8168164 1.0
string_ppi_cooccurence raw_sim 2858x2858 8168164 58344 0.007142853644956198
string_ppi_cooccurence rf 2858x2858 8168164 1409490 0.17255897408524118
string_ppi ct 18476x18476 341362576 341362576 1.0
string_ppi_database ct 10713x10713 114768369 114768099 0.9999976474354184
string_ppi_database el 10713x10713 114768369 109307927 0.9524220650029452
string_ppi_database ka 10713x10713 114768369 696547 0.006069154820872291
string_ppi_database node2vec 10713x10713 114768369 114768369 1.0
string_ppi_database raw_sim 10713x10713 114768369 685840 0.005975862565407722
string_ppi_database rf 10713x10713 114768369 109307927 0.9524220650029452
string_ppi el 18476x18476 341362576 341362576 1.0
string_ppi_exp ct 17248x17248 297493504 297493504 1.0
string_ppi_exp el 17248x17248 297493504 297183146 0.9989567570524162
string_ppi_exp ka 17248x17248 297493504 4327866 0.014547766394253772
string_ppi_exp node2vec 17248x17248 297493504 297493504 1.0
string_ppi_exp rf 17248x17248 297493504 297183146 0.9989567570524162
string_ppi_experimental ct 17248x17248 297493504 297493504 1.0
string_ppi_experimental el 17248x17248 297493504 297183146 0.9989567570524162
string_ppi_experimental ka 17248x17248 297493504 4327866 0.014547766394253772
string_ppi_experimental node2vec 17248x17248 297493504 297493504 1.0
string_ppi_experimental raw_sim 17248x17248 297493504 4310629 0.014489825633301895
string_ppi_experimental rf 17248x17248 297493504 297183146 0.9989567570524162
string_ppi_fusion ct 5970x5970 35640900 35639688 0.9999659941247275
string_ppi_fusion el 5970x5970 35640900 15344240 0.43052335939889286
string_ppi_fusion ka 5970x5970 35640900 28186 0.0007908330036559122
string_ppi_fusion node2vec 5970x5970 35640900 35640900 1.0
string_ppi_fusion raw_sim 5970x5970 35640900 22217 0.0006233568737040872
string_ppi_fusion rf 5970x5970 35640900 15344240 0.43052335939889286
string_ppi ka 18476x18476 341362576 11290078 0.033073566916134355
string_ppi_neighborhood ct 3891x3891 15139881 15139881 1.0
string_ppi_neighborhood el 3891x3891 15139881 15139881 1.0
string_ppi_neighborhood ka 3891x3891 15139881 363637 0.024018484689542804
string_ppi_neighborhood node2vec 3891x3891 15139881 15139881 1.0
string_ppi_neighborhood raw_sim 3891x3891 15139881 359746 0.02376148134849937
string_ppi_neighborhood rf 3891x3891 15139881 15139881 1.0
string_ppi node2vec 18476x18476 341362576 341362576 1.0
string_ppi raw_sim 18476x18476 341362576 11271627 0.03301951588272523
string_ppi rf 18476x18476 341362576 341362576 1.0
string_ppi_textmining ct 18441x18441 340070481 340070481 1.0
string_ppi_textmining el 18441x18441 340070481 340070481 1.0
string_ppi_textmining ka 18441x18441 340070481 9638097 0.028341469014477618
string_ppi_textmining node2vec 18441x18441 340070481 340070481 1.0
string_ppi_textmining raw_sim 18441x18441 340070481 9619658 0.028287247901413706
string_ppi_textmining rf 18441x18441 340070481 340070481 1.0

Table 3. Integrated Embedding Matrixes

Integration Kernel Matrix_Dimensions Matrix_Elements Matrix_Elements_Non_Zero Matrix_Non_Zero_Density
geometric_mean ct 30023x30023 901380529 793518737 0.8803371178655669
geometric_mean el 30023x30023 901380529 488416499 0.5418538378478774
geometric_mean ka 30023x30023 901380529 24617803 0.02731122118569748
geometric_mean node2vec 30023x30023 901380529 793747191 0.8805905668725622
geometric_mean raw_sim 30023x30023 901380529 24588057 0.02727822069473613
geometric_mean rf 30023x30023 901380529 488416499 0.5418538378478774
integration_mean_by_presence ct 30023x30023 901380529 793525165 0.8803442491489629
integration_mean_by_presence el 30023x30023 901380529 568044367 0.6301937402954485
integration_mean_by_presence ka 30023x30023 901380529 268650869 0.29804378989420127
integration_mean_by_presence node2vec 30023x30023 901380529 793747189 0.8805905646537435
integration_mean_by_presence raw_sim 30023x30023 901380529 268638179 0.2980297114893637
integration_mean_by_presence rf 30023x30023 901380529 568044367 0.6301937402954485
max ct 30023x30023 901380529 793318325 0.8801147789159754
max el 30023x30023 901380529 568044367 0.6301937402954485
max ka 30023x30023 901380529 268650869 0.29804378989420127
max node2vec 30023x30023 901380529 793747191 0.8805905668725622
max raw_sim 30023x30023 901380529 268638179 0.2980297114893637
max rf 30023x30023 901380529 568044367 0.6301937402954485
mean ct 30023x30023 901380529 793525165 0.8803442491489629
mean el 30023x30023 901380529 568044367 0.6301937402954485
mean ka 30023x30023 901380529 268650869 0.29804378989420127
mean node2vec 30023x30023 901380529 793747189 0.8805905646537435
mean raw_sim 30023x30023 901380529 268638179 0.2980297114893637
mean rf 30023x30023 901380529 568044367 0.6301937402954485
median ct 30023x30023 901380529 793508477 0.8803257353254854
median el 30023x30023 901380529 567907135 0.6300414938295167
median ka 30023x30023 901380529 55920597 0.062038833989501455
median node2vec 30023x30023 901380529 793747189 0.8805905646537435
median raw_sim 30023x30023 901380529 55892579 0.062007750557922243
median rf 30023x30023 901380529 567907135 0.6300414938295167

Weight values